Azure Schema Registry Apache Avro client library for .NET - version 1.0.1

  • Article

Azure Schema Registry is a schema repository service hosted by Azure Event Hubs, providing schema storage, versioning, and management. This package provides an Avro serializer capable of serializing and deserializing payloads containing Schema Registry schema identifiers and Avro-serialized data.

Getting started

Install the package

Install the Azure Schema Registry Apache Avro library for .NET with NuGet:

dotnet add package Microsoft.Azure.Data.SchemaRegistry.ApacheAvro

Prerequisites

If you need to create an Event Hubs namespace, you can use the Azure Portal or Azure PowerShell.

You can use Azure PowerShell to create the Event Hubs namespace with the following command:

New-AzEventHubNamespace -ResourceGroupName myResourceGroup -NamespaceName namespace_name -Location eastus

Authenticate the client

In order to interact with the Azure Schema Registry service, you'll need to create an instance of the Schema Registry Client class. To create this client, you'll need Azure resource credentials and the Event Hubs namespace hostname.

Get credentials

To acquire authenticated credentials and start interacting with Azure resources, please see the quickstart guide here.

Get Event Hubs namespace hostname

The simplest way is to use the Azure portal and navigate to your Event Hubs namespace. From the Overview tab, you'll see Host name. Copy the value from this field.

Create SchemaRegistryClient

Once you have the Azure resource credentials and the Event Hubs namespace hostname, you can create the SchemaRegistryClient. You'll also need the Azure.Identity package to create the credential.

// Create a new SchemaRegistry client using the default credential from Azure.Identity using environment variables previously set,
// including AZURE_CLIENT_ID, AZURE_CLIENT_SECRET, and AZURE_TENANT_ID.
// For more information on Azure.Identity usage, see: https://github.com/Azure/azure-sdk-for-net/blob/Microsoft.Azure.Data.SchemaRegistry.ApacheAvro_1.0.1/sdk/identity/Azure.Identity/README.md
var schemaRegistryClient = new SchemaRegistryClient(fullyQualifiedNamespace: fullyQualifiedNamespace, credential: new DefaultAzureCredential());

Key concepts

Serializer

This library provides a serializer, SchemaRegistryAvroSerializer, that interacts with EventData events. The SchemaRegistryAvroSerializer utilizes a SchemaRegistryClient to enrich the EventData events with the schema ID for the schema used to serialize the data.

This serializer requires the Apache Avro library. The payload types accepted by this serializer include GenericRecord and ISpecificRecord.

Examples

The following shows examples of what is available through the SchemaRegistryAvroSerializer. There are both sync and async methods available for these operations. These examples use a generated Apache Avro class Employee.cs created using this schema:

{
   "type" : "record",
    "namespace" : "TestSchema",
    "name" : "Employee",
    "fields" : [
        { "name" : "Name" , "type" : "string" },
        { "name" : "Age", "type" : "int" }
    ]
}

Details on generating a class using the Apache Avro library can be found in the Avro C# Documentation.

Serialize and deserialize data using the Event Hub EventData model

In order to serialize an EventData instance with Avro information, you can do the following:

var serializer = new SchemaRegistryAvroSerializer(client, groupName, new SchemaRegistryAvroSerializerOptions { AutoRegisterSchemas = true });

var employee = new Employee { Age = 42, Name = "Caketown" };
EventData eventData = (EventData) await serializer.SerializeAsync(employee, messageType: typeof(EventData));

// the schema Id will be included as a parameter of the content type
Console.WriteLine(eventData.ContentType);

// the serialized Avro data will be stored in the EventBody
Console.WriteLine(eventData.EventBody);

// construct a publisher and publish the events to our event hub
var fullyQualifiedNamespace = "<< FULLY-QUALIFIED EVENT HUBS NAMESPACE (like something.servicebus.windows.net) >>";
var eventHubName = "<< NAME OF THE EVENT HUB >>";
var credential = new DefaultAzureCredential();

// It is recommended that you cache the Event Hubs clients for the lifetime of your
// application, closing or disposing when application ends.  This example disposes
// after the immediate scope for simplicity.

await using var producer = new EventHubProducerClient(fullyQualifiedNamespace, eventHubName, credential);
await producer.SendAsync(new EventData[] { eventData });

To deserialize an EventData event that you are consuming:

// construct a consumer and consume the event from our event hub

// It is recommended that you cache the Event Hubs clients for the lifetime of your
// application, closing or disposing when application ends.  This example disposes
// after the immediate scope for simplicity.

await using var consumer = new EventHubConsumerClient(EventHubConsumerClient.DefaultConsumerGroupName, fullyQualifiedNamespace, eventHubName, credential);
await foreach (PartitionEvent receivedEvent in consumer.ReadEventsAsync())
{
    Employee deserialized = (Employee) await serializer.DeserializeAsync(eventData, typeof(Employee));
    Console.WriteLine(deserialized.Age);
    Console.WriteLine(deserialized.Name);
    break;
}

You can also use generic methods to serialize and deserialize the data. This may be more convenient if you are not building a library on top of the Avro serializer, as you won't have to worry about the virality of generics:

var serializer = new SchemaRegistryAvroSerializer(client, groupName, new SchemaRegistryAvroSerializerOptions { AutoRegisterSchemas = true });

var employee = new Employee { Age = 42, Name = "Caketown" };
EventData eventData = await serializer.SerializeAsync<EventData, Employee>(employee);

// the schema Id will be included as a parameter of the content type
Console.WriteLine(eventData.ContentType);

// the serialized Avro data will be stored in the EventBody
Console.WriteLine(eventData.EventBody);

Similarly, to deserialize:

Employee deserialized = await serializer.DeserializeAsync<Employee>(eventData);
Console.WriteLine(deserialized.Age);
Console.WriteLine(deserialized.Name);

Serialize and deserialize data using MessageContent directly

It is also possible to serialize and deserialize using MessageContent. Use this option if you are not integrating with any of the messaging libraries that work with MessageContent.

var serializer = new SchemaRegistryAvroSerializer(client, groupName, new SchemaRegistryAvroSerializerOptions { AutoRegisterSchemas = true });
MessageContent content = await serializer.SerializeAsync<MessageContent, Employee>(employee);

Employee deserializedEmployee = await serializer.DeserializeAsync<Employee>(content);

Troubleshooting

If you encounter errors when communicating with the Schema Registry service, these errors will be thrown as a RequestFailedException. The serializer will only communicate with the service the first time it encounters a schema (when serializing) or a schema ID (when deserializing). Any errors related to invalid Content-Types will be thrown as a FormatException. Errors related to invalid schemas will be thrown as an Exception, and the InnerException property will contain the underlying exception that was thrown from the Apache Avro library. This type of error would typically be caught during testing and should not be handled in code. Any errors related to incompatible schemas will be thrown as an Exception with the InnerException property set to the underlying exception from the Apache Avro library.

Next steps

See Azure Schema Registry for additional information.

Contributing

This project welcomes contributions and suggestions. Most contributions require you to agree to a Contributor License Agreement (CLA) declaring that you have the right to, and actually do, grant us the rights to use your contribution. For details, visit cla.microsoft.com.

When you submit a pull request, a CLA-bot will automatically determine whether you need to provide a CLA and decorate the PR appropriately (e.g., label, comment). Simply follow the instructions provided by the bot. You will only need to do this once across all repos using our CLA.

This project has adopted the Microsoft Open Source Code of Conduct. For more information see the Code of Conduct FAQ or contact opencode@microsoft.com with any additional questions or comments.

Impressions